REVUE BRYOLOGIQUE. u 
of Jungermania, namely with the subgenus Sphenolobus 
Lind. (— Jg. minuta Schrad., Jg. saxicola Schrad., etc.) 
which it is difficult to keep separate from Diplophyllum Dur. 
(= Jg. albicans, etc.) ; and the minuter species, by their 
habit and pectinate foliage, show some affinity with the small- 
leaved acrocarpous Cephaloziæ (C'. divaricata, exiliflora, ete. ). 
The frontal ftattening of the stem (with its leaves), and 
especially of the perianth , is a character that sometimes runs 
through large groups of genera, e. g. Lejeunea and Frullania 
(with all their subgenera), Radula, Scapania, Cephalozia, etc. 
The lateral flatiening occurs in much fewer species, but is 
equally constant throughout any genus where it exists, e. g. 
in Plagiochila, Lophocolea, Leioscyphus, Southbya, etc., so that 
it cannot but be regarded as an important dislinction, suffi- 
cient of itself (as I conceive) to separate Nardia (or Alicularia) 
from Marsupella (or Sarcoscyphus ). 
As above-staled, the normal and perfectly-developed pe- 
rianth of Marsupella is tubular — often broadly-oval — and 
is in some species closed at the constrictéd apex, in others 
open. But even in typical species, incomplele perianths are 
of frequent occurrence , owing to the anthophyls (as we may 
be allowed to call the leaves, or bracts — interior to and later 
developed than the outer bracts, or involucre proper — whose 
union constitutes the monophyllous perianth, or colesule ) 
having remained more or less disunited ; so that in some 
_cases, the upper portion, instead of forming a complete dome, 
is cucullate, wanting one side of the hemispkere ; or the two 
(or three) anthophyls are connected only at the base, free 
at the apex and often bilobed — like the ordinary stem-leaves 
of which they are but modifications; or they remain free 
from the very base, or at least are connate-on only one margin ; 
or finally one or other of them may be quite obsolete, and 
only a single small leaf remain Lo represent the perianth. 
[All these variations coexist in the flowers of a single species, 
to be described below as M. ustulata.] Even the adhesion of 
the perianth in its lower balf Lo the involuere sometimes di- 
limited to one side only of 
